A great deal of interest has been focused on nitride-based light emitting diodes and lasers that emit light in the blue and deep ultraviolet (UV) wavelengths. These devices may be capable of being incorporated into various applications, including solid-state lighting, biochemical detection, high-density data storage, and the like. However, to date, the performance of nitride-based light emitting diodes and lasers quickly worsens as the radiation wavelength is reduced into the ultraviolet range.
A modern light emitting diode (LED) consists of three major components: an electron supply layer (e.g., an n-type semiconductor) and a hole supply layer (e.g., a p-type semiconductor), with a light generating structure between them. The relatively poor efficiency with which light generated by the light generating structure is a significant barrier to improving the performance of the device in generating light having the shorter wavelengths. Such efficiency is limited by a large difference between the mobilities of electrons and holes. Since electrons are more mobile than holes, the electrons travel more quickly than holes.
To address this situation, some approaches incorporate an electron blocking layer between the light generating structure and a p-type contact layer. The electron blocking layer slows down electrons and allows for a more efficient radiative recombination. However, the electron blocking layer also increases the series resistance of the device and, to a certain extent, provides a barrier for holes as well. Many approaches incorporate multiple quantum wells into the light generating structure to increase the concentration of electron-hole pairs. However, these approaches still fail to provide a solution that efficiently generates light in the shorter wavelengths. Since an amount of nonradiative recombination of electrons and holes is determined by dislocations, many approaches seek to improve the quality of the materials used in the device. Nevertheless, the efficiency of deep UV LEDs remains low.
Another difficulty in developing a UV LED is a deficient hole injection. To date, Magnesium (Mg) is the most successful acceptor, and is therefore commonly used for p-type Gallium (Ga) Nitride (N) layers. The room-temperature activation energy for such a layer can be as high as two-hundred fifty milli-electron Volts (meV), and increases roughly linearly with the Aluminum (Al) molar fraction in AlGaN alloys. However, a large acceptor activation energy results in a deficient hole injection. This is particularly true for a deeper UV LED, in which a higher Al molar fraction is required.
Various approaches seek to enhance the conductivity for a p-type Mg-doped AlGaN layer. In one approach, a Mg-doped AlGaN/GaN short period superlattice (SPSL), such as a Mg-doped AlGaN/GaN SPSL in 340-350 nm UV LED growth, has been used in place of the layer. In this case, a period of the superlattice is sufficiently small (e.g., below four nanometers) so that the effect of the polarization fields on the minibands in the SPSL is negligible. As a result, a vertical conduction of the p-type SPSL is not degraded by the polarization fields.
Another approach uses a Mg-doped AlGaN/GaN large period superlattice (LPSL). In this case, with a period larger than fifteen nm, a valence band discontinuity and the polarization fields can enhance the ionization of the acceptors in the AlGaN barriers and transfer holes into GaN wells. However, the large period inhibits the wavefunction coupling between neighboring wells, which greatly reduces the vertical conductivity. As a result, this LPSL approach can only enhance lateral horizontal p-conductivity. To date, no known approach has successfully used a p-type LPSL for a deep UV LED.
Yet another approach uses a p-GaN/p-AlGaN single heterostructure to accumulate holes at the interface. The mechanism of this approach is similar to the LPSL approach. However, since the p-GaN/p-AlGaN single heterostructure only includes one barrier for hole transportation, the vertical conductivity can be greatly enhanced due to the high-density hole accumulation at the interface, field assisted tunneling, and thermal emission. Several UV LEDs have been proposed that incorporate this approach, and have achieved reasonably good output powers. However, it remains desirable to improve the output power and/or efficiency of UV LEDs.